This project addresses the following hypotheses: 1. Virus will be spontaneously produced after stopping early ART from the earliest-established ?reservoir cells? in mucosal tissues (which Project 1 aims to identify). Candidates for the earliest-established reservoir cells in mucosal tissues (follicular helper T (Tfh), central memory T (Tcm), and Th17 cells) will each be studied ex vivo. T cell latency/reactivation models already described in the literature will be adapted and new models developed that use explanted human and macaque colon mucosa, as well as human tonsils. The former aims to study Th17, and the latter Tfh, cells as potential early reservoirs that may initiate rebound after ART. We will also pilot humanized DRAG mice as a model of female reproductive tract (FRT) mucosal Tfh reservoirs, and determine whether the tonsil explants or humanized mice are preferable for further study of Tfh cell-related mechanisms and interventions (in aims 2 and 3). Macaques with ART-suppression starting at day 3 after mucosal SIV infection, and continuing for 6 months before stopping, will also be studied in close collaboration with project 1. An observational study of HIV-infected subjects before and after pausing >1 year-long suppressive ART will be designed and performed (?Clinical Observation of Rebound after pausing ART?, CORA). 2. mTOR-mediated metabolic reprogramming triggered by T cell activation contributes to viremia rebound off-ART. Cellular mechanisms downstream of mTOR activity will be studied using catalytic mTOR inhibitors in ex vivo tissue and animal models, including in SIV-infected macaques suppressed by ART for 6 months before stopping it. Effects on both virus production from reservoir cells after early ART stops, and target cell susceptibility to virus, will be characterized. 3. Limited or absent virion A3G in virions initially-produced after stopping early ART contribute to the pathogenic cascade to viremia rebound. Model systems (ex vivo and/or humanized DRAG mice) will be used to test whether virus reservoir T cells have low A3G levels even before Vif is expressed (contributing to virus infectivity), and if A3G- boosting tool compounds decrease infectivity of virions produced from T cell reservoirs. We will assess if A3G- mediated decreased virus spread adds to mTOR inhibitor-mediated decreases in uninfected target T cell susceptibility to infection, and if changes in A3G's cytoplasmic distribution caused by mTOR inhibition enhance A3G booster-mediated increases in A3G virion packaging.